Godfrey has been lightly using JiTR to pull media reports on coffee growers in Costa Rica. For now, Godfrey prefers the system for data collection, retention and organization. It is these things he thinks will benefit him in the long run, hoping to have, in the end, a comprehensive and easily explored collection of annotated notes. Using JiTR's Knowledge Manager, he has created a spider that searches through his common sources and fetches new stories. When a article come a in, he is notified via email. Logging in, he proceeds to read over the new item, adding his notes about the content and organizing it through relevant labels. Labeling has hit a note with Godfrey, because when he remembers to do it, he can easily browse through items that he needs at a particular time. Godfrey is beginning to warm to the system, and decides to add his older research to the repository, manually specifying its chronological place in the repository. Nevertheless, he still does not fully trust the web, and religiously exports offline backups of his repositories.

As he is preparing for a meeting with his colleagues, Godfrey chooses a selection of items compiled in the last few weeks and exports them to a single, easily-accessible page. He sends the link out prior to the meeting to provide background information. For the meeting, he prints a bibliography with his notes for each item. An alert appears, informing Godfrey that one of his articles is missing an author. He clicks the provided link, and a few seconds later, he has added the proper information for JiTR.

---

NOTES: "For the meeting, he prints a bibliography with his notes for each item" : As Susan noted, 'bibliography' is not the right word here, as it suggests adding a complicated referencing tool. The intention was to have JiTR organize the metadata it knows (author, data, data accessed, title, url) in a familiar way.

Point-Form

First Encounter

• Godfrey gets a basic account created automatically for him.
• He creates a repository for "Coffee Research". (Image 1)
• He has some web pages of materials bookmarked so he tries adding them to his repository by using the Simple Acquire tool that, given a URL, will add the item.(Image 2)
• After adding dozen items he tries some of the analytical tools to see what they will do.
  • He tries the word cloud and list words. This suggests some ideas for keywords for his tags and he goes through and starts systematically tagging items by creating categories that he can check off.
  • He tries the cleaner and runs that on each item to see whether it leaves the text he wants. He finds he still ends up deleting text left after the cleaner to have only the key passages. He is satisfied that he can always retrieve the original version. (Images 3,4)
• Having done all this work he wants to see if he can save the materials to his hard drive. He doesn't trust web research environments so he exports his repository with the notes and tags. He is satisfied that he can keep the materials on his PC if something goes wrong with the repository. (Image 5)

Second Encounter

• Godfrey now wants to see what the more advanced tools do. He sets up a spider to start from his common source sites. (Image 1)
  • New items are automatically fetched
• Godfrey receives an email every day summarizing the new items spidered.
• After a week he decides to see what was retrieved and clean up the repository.
  • He reads each new item and decides whether to keep or discard it. With one click he can discard items. (Image 2)
  • If he keeps it he uses the automatic cleaner to remove the HTML (Image 2)
  • On the same page, he then gives items category labels (using categories developed before and adding some) (Image 2)
  • He adds a short note to the ones that are important. (Image 2, Image 3)
  • Godfrey also makes a point to add a readable label, because that helps him scan his list of items.
• he starts adding older research, for the organizational benefits (Image 4)
• Godfrey is now satisfied that this can help him and he again backs up his repository to a tagged text file he can consult. He chooses an HTML format so he can easily browse his items on his PC.

Third Encounter

• Godfrey wants to share his repository with a colleague at a meeting. He needs to first get old stuff in.
• He enters old items that he had in his files and which are not available on the web to acquire. Some he has to type and some he can copy from notes typed into his PC.
• To make a selection of items visible to his colleague he tries goes through the repository and marks important items as public. He notes the URL for the public items from this repository which he sends to his colleague.
• As a backup Godfrey also exports a single HTML page, but without his notes to e-mail his colleague
• For his meeting, he also prints out a report about the repository which summarizes how many items and includes his notes.
• At the meeting he and his colleague decide to collaborate and use the JiTR repository. Now Godfrey has to figure out how to give Jane access.

First Encounter

Second Encounter

New taxes on coffee trade in Costa Rica has growers setting up protests against the government, claiming that, if anything, they should be subsidized, not taxed. Godfrey start tagging articles on the topic with the term "Coffee Tax Conflict". In the notes for each article, Godfrey adds a small summary of the article and his analysis of its biases.

Before trying to analyze the bigger picture, Godfrey runs the "Count Words" process that JiTR offers to him through its TAPOR connection. Running the process on the "Coffee Tax Conflict" tag, organized by chronology, he is given bar graphs that show the frequency of each word throughout the larger collection. One thing in particular catches Godfrey's eye: while mentions of the president by name go down over time, the word "spokesperson" goes up. Also, the use of the word "tax" decreases. Godfrey forms a hypothesis as to the change in reporting over time, and skimming through the articles, confirms his guess. It appears that, over time, the media began to broaden their focus onto the larger issue of government confidence, and Godfrey suspects that less of an official voice from the government is what causes the media to stray from the primary issue. He writes this observation in the "Notes" section of the "Coffee Tax Conflict" tag, and a small icon appears beside the term in his tag list, reminding him that there is a note included.

Point-Form

Fourth Encounter

• Godfrey and Jane have been adding items manualy and continuing to run the spider.
• Godfrey finds a couple of sites that have a lot of materials so he uses a crawler to acquire the entire sites.
• Godfrey and Jane agree on a list of tag categories to help them manage the sub-issues. Jane goes through retagging the items. She finds she can use the bulk tagging tool to change tags and she can search the repository for keywords to suggest which items should be tagged a certain way.
• When he reads each new article, Godfrey adds pointer notes about it in the "Notes Section".
• Ability to run TAPOR Tools is useful when Godfrey runs "Count Words" on a chronological collection of a specific tag. This lets him see word use over time
• Notes are also allowed for a specific tag.

• Since notes can be added to entries, tags, and the collection at large, small icons are used to easily show when there is a note attached to something.

Sydney needs to identify and explore a series of colloquial terms unique to the Franco-Ontarian population.

He assembles a custom web spider routine using JiTR's drag and drop spiderBuilder. He runs this spider to construct a collection of articles drawn from Franco-Ontarian sources. He then applies a recipe he found in the TAPoR Portal to identify colloquial word usage in bodies of text. The TAPoR tools are provided as a plug-in to the JiTR environment and this enables him to quickly isolate word groups meeting his needs. Sidney's results are added to his repository as a separate text. These isolated phrases then serve as a target for additional analysis to discover patterns in their usage with additional tools available from the JiTR dashboard.

• Sydney is studying Franco-Ontarian colloquial terms
• sets up a webspider using the drag and drop spiderBuilder
• uses a TAPoR recipe to identify colloquial word usage
• results are sent back to JiTR, into the repository

Sydney is collecting French language shareholder information from Canadian companies, in hopes of comparing them to their English counterparts.

To collect the documents, Sydney uses that manual item-add. Since they are all online, however, Sidney does not need to upload them. Rather, he adds the documents by entering the target URLs.

Since the documents collected are in PDF, Sidney uses the PDF-to-Text coversion process to create more tangible items with them. Once he does that, he adds anchor targets to each header of the documents. With this done, he is able to create links between the French and English versions, so he makes each section header link to its counterpart (in the other language).

• Sydney is scollecting French and English Canadian shareholder information
  • he adds the items by pasting in their URLs
• Sydney converts his PDFs to text items
• he adds anchors to section headers (with the items), and makes each header a link to its other-language counterpart

Mandy has been tracking her company in the media as well as, at the same time, keeping an eye out for trademark infringement. It has been time consuming: she searches through newspaper databases, runs google alerts, and constantly tracks the for fifty search results for "Maplesoft". When a colleague suggests "that cool new JiTR thing", she decides to give it a try.

Upon first visiting the site (Image 1), the main page has various informations, including a video explanation of the system's functionality and blurbs showing various creative ways that people have been using the site(Image 2). One of these short blurbs outlines a "commercial user", so she clicks on the accompanying link for more information, and is presented a more detailed page of how commercial users can use JiTR. Finally convinced, she goes through the simple sign-up process (Image 3). Upon first log-in, there is an example repository in her account (Image 4), with items that further explain how it works.

Scenario 1 Point-Form

• Mandy hears about JiTR, but doesn't yet fully grasp it
• She goes to the page to test it out, and finds a lot of information about it
• The information looks at functionality, but also, in a more practical sense, specific uses by different types of people. There is also a video walkthrough
• Mandy clicks on one of the blurbs that is similar to her uses, and finds out more information
• convinced, Mandy signs up
• upon first login, there is a "tutorial" repository

Scenario 1 Wireframe

Mandy sets up two repositories to help her track mentions of her company in the media, an ongoing task in her position. The first repository looks at general mentions online. First, she follows the steps of the Repository Wizard, where she names the repository ("General Maple Mentions"), sets up the style of item collection ("web spider") and sets the verbosity of site instructions ("very"). She sets the web spider to search for new instances of Waterloo Software or their product, "Maple 11" being mentioned online. So as to receive less irrelevant information, the Wizard suggests that Mandy populate a list of relevance keywords, such as "algebra","tool", and "programming", which then allows her to set a threshold of probable relevance. After asking what she is using the tool for, the Wizard suggests that Mandy organize her items using priority tags (e.g. "1" for most credible source, "3" for least important source).

After completing her initial repository setup, Mandy returns to the Knowledge Manager to add addition ways of collecting items. She sets up a tool to monitor the changes of the "Waterloo Maple Software" Wikipedia pages as well as blog search mentions, categorizing accordingly the items obtained from these. If the web spider,wiki-tracker, or blog-watcher overlap, JiTR's instructions assure Mandy that the system won't put the same entry in twice.

After a few days of testing JiTR, Mandy creates her second repository. This one is of major news and business news mentions. Shes sets it up similar to her first, except that the Wizard offers her a template that includes a web search with a "major media" filter list and a search tool for Lexis-Nexis. Since JiTR stores all the circumstances (metadata) of an item's amassment, the filtered web search of the second repository is able to scan the first repository, and gather any items that would have been pulled in had the second repository's search been running earlier.

---

NOTES: As Geoffrey noted, why would she be using two repositories? A good (and I would add 'well understood') tagging/labelling system would make it unnecessary.

Scenario 2 Point-Form

• Mandy has two repositories: one of general mentions of the company or company software, and the other of major mentions
• To setup her repository, she starts the wizard
• Mandy sets up a search engine spider, then goes back and sets up a blog tracker and a Wikipedia page change tracker
• The wizard suggests a way to tag her items (priorities)
• the search engine spider allows her to put in other keywords, to help its relevance threshold
• the system is good at hand-holding Mandy through the process
• Mandy's second repository is similar to the first, but the system offers a template, which searches major news sites as well as Lexis-Nexis

Scenario 1 Narrative

When he hears about a Russian blogger causing controversy by speaking out against Russian involvement of Latvian affairs, G.G. starts a repository to track him. He is still unsure of what his end goal is, but knows that this story may well informing his research. What JiTR offers to him is two things: the ability to save posts in case they get taken down, and a malleability of working with and sort posts. He sets up an aggregate function to pull posts from the blog. In the options, Graydon sets the history feature, to update the item daily if it has changed on the site. Graydon worries that the blogger may suddenly turn coy, and silently start rewriting old posts to be less inflammatory. If that were to happen, which it thus far has not, Graydon would be able to work through a timeline of his item. How the function has proved useful is by providing a snapshot of a specific time. When there is an online mention of any of the blogger's posts, Graydon can see exactly where the comments were when the mention was published, and can retrace how it may have affected the bloggers comments.

Graydon tracks mentions of the blog, and chooses to do it within the same repository (for an easy view of the entire situation as it unfolds). Since the two functions are mixed, he sets the items that are pulled from the blog to be highlighted in bright yellow, so as to stand out. He also likes the timeline graph function, which shows Graydon the frequency of online mentions at any given time, and lets him track how much activity there is as a result of every new post.

Scenario 1 Point-Form

• Graydon is tracking a Latvian blogger. His end result is analysis, but he does not know how. For know, it's just tracking
• he sets up an ongoing aggregator, which takes the syndicated blog feed and pulls it into JiTR
• he turns on the history feature, which tracks for post changes and updates repository items with a new version number
• Graydon tracks mentions of the blog also, within the same repository
• Graydon sets the colour of the Latvian blogger's items to have a highlighted yellow background